Okay, so I've played in a few games, mostly under one DM. I've been working on my own campaign setting, but it just seems like there is so much to work on before it is even ready to try running. Where do you guys start? What parts of the world do you feel are "musts" for fleshing out first, and what do you play be ear as the game continues? I've been trying for a while, and so much gets thrown in the way that I don't know if I'd ever really be able to finish in this lifetime.

~SE++

The first thing I do is figure out the setting's over-arching theme - something like Dark Sun's post-apocalyptic struggle for survival or Ravenloft's darkness-at-every-turn set-up. The core identity of a setting, to me, should immediately leap out and say what kind of games can be run well given the landscape you've envisioned.

Once the big question of "What is this world going to be about?" has been answered, I like to start with a slew of other important questions. Typically the first is "What kind of peoples and races exist in this world?". Is this a setting where humans are the only sapient race and dominate, and monsters lurk on the fringe of the world, like Westeros? Are the setting's inhabitants closer to the Tolkein-inspired Core Four, or are monstrous races like dragonborn and tieflings able to fit in somewhere? How have the peoples divided in the world - have nations formed in which one race reigns supreme, or do multiple races band together to form a national identity?

From there, I focus on things like the gods (whether they exist with absolute certainty a la Forgotten Realms or whether they "exist" in the hearts of their faithful like in Eberron), magic (is it common or rare?; does it have a legacy of good or ill?), and nations/geography (how big is the setting?; is the setting one area, or one area amidst many neighboring lands?). I give special attention to the area in which I intend to start a game. Race, magic level, and gods are all useful for the players to know, so it's useful to figure these things out to some level towards the beginning of the process.

Of course, the trick is to only give detail to the things that might actually come up in a game, and personally, my homebrew world arose around the first ideas I had for a campaign. I knew when I sat down to take notes that I wanted a specific game, and from that idea, I was able to slowly build the world from the inside out. The things I felt would be most relevant were built at the center, and as my ideas expanded, so did the scope of the world, but those later details that emerged were simply the result of the game progressing.

Example: If my campaign is going to be set in a monarchy, then I'll make sure to hammer out who is in charge, what are the big settlements, what is the culture like, and what are the big events in its history. However, if I know that nothing is ever going to come of this monarchy bordering a republic, then why devote time to detailing that republic when things can come on the fly?

I'd say don't be afraid to just focus your efforts on the area in which you expect to launch a campaign, whether it is an island surrounded by monsters, a city rife with political corruption, or a backwater village threatened by undead. As long as you have a few ideas of the greater world that surrounds that immediate area, you don't need anything more; worldbuilding is a fun process, but don't worry about the places and names that aren't going to play a role in your game. They can come later, when you can breathe easy knowing you have the important pieces on the board.

Thanks for all the great info!

I guess my idea was a little more broad, in that I wanted a WORLD. Not just a campaign setting, but a full world where adventurers could go freely from city to city without me needing to stop the campaign for a week to flesh out the next city. I wanted the world to have loads of different areas to explore, different cultures to encounter and a setting where anything can and will occur if the players want it to. I wanted to have everything from cities created, to gods named, to custom creatures outside of the bestiaries, and everything in between. I dunno, perhaps that is too lofty a goal, but in my mind, I need to know how the whole world will react to an event that triggers a campaign, not just a single town.

~SE++

The way I do things is similar to what LordManshoon said above:

Think of it like a puzzle... you want it to look like a picture of a bowl of fruit. You dont start off with no basis on what a bowl of fruit looks like, and you don't just start hooking random pieces together to see what fits.

Start with a framework, like the edge pieces of the puzzle:
Whats the main "setting" is it a world torn by war? An uneasy peace? Plagued by the return of the dragons? Once you know your main parameter its easy to start work on some of the smaller things.

Next start working on the easy to find and put together pieces. All the apple pieces are red, grab em and get to work:
Start with a core idea for each category and build upon it. What are the gods? How are the viewed? Whats the basic landscape like? Whats the average person's life like?

Start connecting your 'islands' of completed ideas:
The more small things you solve, the more of them you can clump together. Then before you know it you'll have larger and larger chunks of things accomplished.
Its not easy to think of a city-state, and know its story if you dont know things about it. If you've already established how people view the gods, whether they are at war or peace, what their lifestyle is like who their friends and allies are.... suddenly you'd begin to see how that city would react organically in the world you've created around it. Rather than having it sit on a "blank page" and try and brainstorm ideas.

I start with just one city/town and go from there. In fact once I find my notes I plan to post one of my cities.

I generally do it like Trapped posted. I'll start with a town or city, sometimes - rarely - a region. I have an overarching theme of how I want the campaign to go, but I also have the overarching theme of how I want the first couple of adventures to go. And I give those priority.

After I have my starting location, I give the basic info about that place to my players. Then they build their characters, based on what I've established. Once I get their characters - particularly their background information - I can expand just a little bit. Maybe Joe didn't want to be from the town I created, so he said he was from the village further up-river. Bam! Now I have a new community a hundred or so miles away, and I have at least one player who has a personal interest in that community.

Then I just play. As the adventures pile up, you'll have more and more locations. Add them to your main map where it makes sense to have them. And build from there.

Then, your next campaign can start on the edge of where your previous one ended. The players can go back where they've been before and it'll be consistent to them, or they can go and explore further. If they do the former, you have all your notes and such from previously to keep things moving; if they do the latter, you get to do more building.

Now... all that said, I really no longer have the time to do world-building I'm not getting paid for. But I also have a general dislike for fully pre-built worlds. Particularly getting saddled with all the history since I have players who could serve as historians for Greyhawk and FR; they know those worlds' histories better than their own. So what I do there is steal the map, the deities, and most of the current political climate in cities and regions. And I make it real clear that the history they know may or may not apply - don't assume because the book says something happened 100 years ago that it happened in my version of the world. And even when I do this, it goes back to the above - for my game, I only worry about one location. And I build from there as needed.

I think it is possible to do this, but you still need to figure out what your goal or intended audience is. The real world has the luxury of just being, but your fictional world is a setting for something.

I think an issue with figuring everything out beforehand though is that it actually makes your world a lot more rigid and less reactive, if you are looking at a D&D game setting.

I agree with what is above, but I will add that I usually have a pretty good idea of what my permitted classes will be. This is important if the region you start has certain classes that are excluded for a reason, or not present elsewhere. I try to be as flexible as I can and allow classes people want, but if it will be too disruptive (ie. wizards in Middle Earth) or too contrary to the setting, it would be good to have that understanding in place when your players pick. Otherwise, I work hard to give them what they ask for, because they have more emotional investment then.

Also, if you add classes beyond the standard, like prestige classes out of 3rd Edition, or d20 books, etc, make sure your players know they are an option.

Since I also make whole worlds, I flesh out geography for all major continents to some degree, and large centers of organization and race location, which kind of speeds up filling in details. A desert empire speaks to certain classes, geography, and adventures naturally versus a feudal kingdom set in the jungle, for example. Paladins will be less likely in the desert, at least in the archetypical form everybody knows. Same thing in a jungle setting. I like to just let details flow from the big picture to help develop the smaller details organically if I don't have something particular in mind.

I always end up starting a town that spins off into a whole world/multiverse before I even decide on a mayor, and then I lose control of everything because it becomes too much work like an hour in. This is why I am usually a player lol.

This. So much this.

But I'll keep trying, and hopefully I'll get to see some games and see what you guys put into your worlds. Maybe it'll inspire me to get going with it.

~SE++

If you want to flesh out everything before hand, you have definitely bitten off more than you can chew. Even creating enough content to get players to level 5 can be daunting with the amount of preparation. (I'm talking 2nd edition here). Episodic encounters are what I use. I create a number of one-of scenarios I can sneak into any adventure to solve problems. So, flesh out the immediate areas for the next few gaming sessions, plus create a few encounters you would use to handle things like players try to go someplace you aren't ready for, like bandits stealing or kidnapping something they want back, which is a great way to lead them, or an NPC they have heard of that has a great/sob story to tell that hopefully leads them to wanting to help.

Generally, if you are having a hard time controlling the players actions, its because they are **** players. It is an INSANE amount of work to flesh out areas completely. Random encounters, still rooted in your world, rule.

Step 1 takes a couple hours. I start with a Geographical Map at a 1:10,000,000 scale or higher. Usually 1 map at that scale is enough, my worlds tend to be smaller than earth, even if there are a number of "continents."

Most of the time it looks like Blue Marble Next Generation (Blue Marble 2004), but with colored pencils. I'll name some oceans, seas, mountain ranges, deserts, island chains, major rivers, etc. I'll name the continents. I'll draw out wind currents and indicate climate ranges (arid, temperate, sub-arctic, etc). I'll identify any major manifestations of magic that are related to the land itself.

It doesn't have to look like a professional map, and you don't have to waste a bunch of time thinking about how you want the landmasses to be shaped. These things form randomly, just make a circle, oval, or rectangle on a sheet of paper and draw squiggly lines in it. Then erase stuff that you don't like. It shouldn't take long. I use 1 sheet of 8.5 x 11 inch paper, sometimes front and back. Usually I draw a Robinson Projection (an oval map).

I know it sounds like a lot, but it isn't. Just draw stuff, the players will like it regardless. No one ever, ever says something like "Wow, your continents are really well formed," or "That's a stupid place for a sea, what were you thinking?!" I mean unless you have a geophysicist in your group.

------------------------------

Step 2 also takes a couple hours. I trace the outline of the continents and land masses on another sheet of paper and draw the Political Map. Use some of your major geographic/geologic/oceanographic/etc features as borders and contested borders. There's no real reason to have more than a couple dozen countries max, your players might actually visit something like 10 kingdoms/countries ever in a very long campaign. Decide on the cultures/races of the major countries, which countries have major conflicts and where those conflicts take place regionally. Draw out some major trade routes. Decide which countries form alliances together and which are part of larger empires.

Mark in some capitol cities and other major civilized destinations. Make some ancient ruins. Decide where technology is at, where art and music are flourishing, where are the major sites of worship or physical manifestations or marks of the deities (if any). This is the map showing things that have *impacted* your world, rather than just its geography. But again, just like the geographic map, don't think too hard. Just drop things down...force yourself to draw lines even if your brain objects for "reasons." Ignore your brain, just draw random countries and stuff.

Use random name generators to make random names for places. Seriously. Maybe name a handful of places yourself if you're really attached to certain names, but otherwise ain't nobody got time for that. There's a lot more things to name on your Political Map than on your Geographic one.

---------------------------------------

Step 3 also takes a couple hours. Build a couple closer up region//city maps of the area the players start in, maybe with some roads and a few towns//villages. A couple maps. Not ten, or twelve, or whatever. Unless you have multiple days//weeks to spend on this, don't try to go all out. These maps should be usable for actual play for the first several sessions of your game.

----------------------------------------

Steps 1 through 3 take a Saturday or a Sunday — maybe a Saturday plus a Sunday tops. But these maps can last you for years, and you can play multiple campaigns on them. When your players want to a explore a new place in your world it's going to be so much easier to get them there, narrate them through their traveling experiences on the way, and build a couple new region maps for the new place when they arrive.

I'm telling you right now, that the way miss_bun has approached world building is the same reason that she fails at it when it spirals out of control.

Don't try to design a Forest by starting at the Trees.

Zoom out, and start at the top. Start at the view looking down on your world from space. Draw that. Make that...it doesn't take long. It doesn't have to be detailed. Worry about that later.

All you're doing is building a mesh outline, basically, of the world. You'll fill in places with more detail later. And you'll be thankful you already have that outline in place.

You can hammer out entire worlds in a day. Just start from above.

Here's some examples from about 15 years ago, so they weren't even colored, that I had scanned in ages ago.

Geographic Map

Political Map

The political map is "clean" because I turned it into a flash animation that the players could mouse over to see continent/country names and statistics and things like that. The dashed lines indicate contested borders or regions of conflict. Dashed lines across water indicate naval conflicts. (One lesson I learned from this particular world is I made waaaay too many countries. The major campaign that actually played on this world saw the players have an actual impact on every continent, and still we really only interacted with a dozen "governments." But at least I already had the options if needed.)

Those two maps have lasted ages, they took a few hours to make by hand and they are really simple examples of what I just wrote about. (I did some back of the envelope calculations, and the surface area of all the land mass on that world is about the same amount of land as Asia — many "worlds" that are custom made are much smaller even than that).

Everything you need for random encounters, events, challenges the players will come across in their travels, and etc, is in those maps. The hand drawn political map talks more about races, capitols, etc.

And then for your detailed stuff, just cross those bridges when the players actually want to get to those places (and reuse, reuse, reuse). You don't need to have every town built up front.

That's a lot of information to work with... maybe I should start with a town. A small one, and see if I can go from there...

~SE++

I usually start with a single town and dungeon and move outward from there. I build as the players move around and when they wish to return to a place I already have the map and the people. All I have to do is throw a new plot in and maybe add an NPC and/or a doorway to a dungeon (sewer grate, a large building they never explored, etc...etc...). If you do that often enough you'll have a pretty large sand box style world to run.

Well, let's see what happens. I started with a fairly fleshed out city, and now it's about to get run here. Wish me some luck!

~SE++